Datasets:

ArXiv:
License:
VISTA-UMI-5K / README.md
tgzzglz666's picture
Update README.md
273f54b verified
metadata
license: apache-2.0
task_categories:
  - robotics
configs:
  - config_name: task_001_press_the_power_button_of_the_power_strip
    default: true
    data_files:
      - split: train
        path: task_001_press_the_power_button_of_the_power_strip/data/*/*.parquet
  - config_name: task_002_transfer_the_calculator_and_then_place_it_on_the_table_to_align
    data_files:
      - split: train
        path: >-
          task_002_transfer_the_calculator_and_then_place_it_on_the_table_to_align/data/*/*.parquet
  - config_name: task_003_the_plug_is_inserted_into_the_power_strip
    data_files:
      - split: train
        path: task_003_the_plug_is_inserted_into_the_power_strip/data/*/*.parquet
  - config_name: task_004_unfold_the_cleaning_cloth_and_wipe_the_table
    data_files:
      - split: train
        path: task_004_unfold_the_cleaning_cloth_and_wipe_the_table/data/*/*.parquet
  - config_name: task_005_scaffold_handover_and_alignment
    data_files:
      - split: train
        path: task_005_scaffold_handover_and_alignment/data/*/*.parquet
  - config_name: task_006_throw_the_napkin_in_the_trash_can
    data_files:
      - split: train
        path: task_006_throw_the_napkin_in_the_trash_can/data/*/*.parquet
  - config_name: task_007_put_the_usb_flash_drive_in_the_storage_cabinet
    data_files:
      - split: train
        path: >-
          task_007_put_the_usb_flash_drive_in_the_storage_cabinet/data/*/*.parquet
  - config_name: task_008_clean_the_sponge_and_put_it_into_the_basin
    data_files:
      - split: train
        path: task_008_clean_the_sponge_and_put_it_into_the_basin/data/*/*.parquet
  - config_name: task_009_throw_away_the_snack_packaging_wipe_the_table
    data_files:
      - split: train
        path: >-
          task_009_throw_away_the_snack_packaging_wipe_the_table/data/*/*.parquet
  - config_name: task_010_put_tableware_into_the_chopstick_cage
    data_files:
      - split: train
        path: task_010_put_tableware_into_the_chopstick_cage/data/*/*.parquet
  - config_name: task_011_put_fruits_into_the_freshkeeping_box
    data_files:
      - split: train
        path: task_011_put_fruits_into_the_freshkeeping_box/data/*/*.parquet
  - config_name: task_012_put_the_mat_on_the_table_and_then_put_the_cup_on_the_mat
    data_files:
      - split: train
        path: >-
          task_012_put_the_mat_on_the_table_and_then_put_the_cup_on_the_mat/data/*/*.parquet
  - config_name: task_013_clean_the_cup_with_a_brush
    data_files:
      - split: train
        path: task_013_clean_the_cup_with_a_brush/data/*/*.parquet
  - config_name: task_014_open_the_lid_and_add_water_to_the_rice_cooker
    data_files:
      - split: train
        path: >-
          task_014_open_the_lid_and_add_water_to_the_rice_cooker/data/*/*.parquet
  - config_name: task_015_put_the_measuring_cup_and_wooden_spoon_on_the_tray
    data_files:
      - split: train
        path: >-
          task_015_put_the_measuring_cup_and_wooden_spoon_on_the_tray/data/*/*.parquet
  - config_name: task_016_wipe_the_countertop_and_place_the_small_pot
    data_files:
      - split: train
        path: task_016_wipe_the_countertop_and_place_the_small_pot/data/*/*.parquet
  - config_name: task_017_put_the_carrot_on_the_board
    data_files:
      - split: train
        path: task_017_put_the_carrot_on_the_board/data/*/*.parquet
  - config_name: task_018_put_the_lid_on_the_cup
    data_files:
      - split: train
        path: task_018_put_the_lid_on_the_cup/data/*/*.parquet
  - config_name: task_019_open_the_drawer_clip_the_eraser_and_put_it_in_the_drawer
    data_files:
      - split: train
        path: >-
          task_019_open_the_drawer_clip_the_eraser_and_put_it_in_the_drawer/data/*/*.parquet
  - config_name: task_020_open_the_drawer_clip_the_stapler_and_put_it_in_the_drawer
    data_files:
      - split: train
        path: >-
          task_020_open_the_drawer_clip_the_stapler_and_put_it_in_the_drawer/data/*/*.parquet
  - config_name: task_021_put_the_ham_sausage_into_the_freshkeeping_box
    data_files:
      - split: train
        path: >-
          task_021_put_the_ham_sausage_into_the_freshkeeping_box/data/*/*.parquet
  - config_name: task_022_the_spatula_goes_into_the_drawer
    data_files:
      - split: train
        path: task_022_the_spatula_goes_into_the_drawer/data/*/*.parquet
  - config_name: task_023_pick_up_the_object_and_put_it_into_the_sock_storage_box
    data_files:
      - split: train
        path: >-
          task_023_pick_up_the_object_and_put_it_into_the_sock_storage_box/data/*/*.parquet
  - config_name: task_024_pick_up_objects_and_put_them_into_the_sock_storage_box
    data_files:
      - split: train
        path: >-
          task_024_pick_up_objects_and_put_them_into_the_sock_storage_box/data/*/*.parquet
  - config_name: task_025_one_hand_picks_up_the_pen_and_the_other_hand_takes_off_the_cap
    data_files:
      - split: train
        path: >-
          task_025_one_hand_picks_up_the_pen_and_the_other_hand_takes_off_the_cap/data/*/*.parquet
  - config_name: task_026_pick_up_the_vase_and_insert_flowers
    data_files:
      - split: train
        path: task_026_pick_up_the_vase_and_insert_flowers/data/*/*.parquet
  - config_name: task_027_put_the_items_into_the_small_pillow_storage_box
    data_files:
      - split: train
        path: >-
          task_027_put_the_items_into_the_small_pillow_storage_box/data/*/*.parquet
  - config_name: task_028_folding_square_towels
    data_files:
      - split: train
        path: task_028_folding_square_towels/data/*/*.parquet
  - config_name: task_029_put_baby_products_into_the_sterilizer
    data_files:
      - split: train
        path: task_029_put_baby_products_into_the_sterilizer/data/*/*.parquet
  - config_name: task_030_put_the_baby_shoes_into_the_container
    data_files:
      - split: train
        path: task_030_put_the_baby_shoes_into_the_container/data/*/*.parquet

VISTA: Vision-Grounded and Physics-Validated Adaptation of UMI Data for VLA Training

Project arXiv Code UMI-VQA-8M VISTA-UMI-5K


Overview

VISTA-UMI-5K is a verified robotic manipulation dataset containing 5,000 UMI trajectories across 30 tabletop manipulation tasks.

All trajectories are high-quality UMI demonstrations validated through our physical-validation pipeline and verified for high-fidelity replay on three representative robot embodiments:

  • RealMan rm7500
  • ARX AC one
  • Galaxea R1 Pro

Each task is organized in LeRobot v3.0 format.

Task List

ID Task Episodes
001 Press the power button of the power strip. 166
002 Transfer the calculator and then place it on the table to align. 166
003 The plug is inserted into the power strip. 166
004 Unfold the cleaning cloth and wipe the table. 166
005 Scaffold handover and alignment. 166
006 Throw the napkin in the trash can. 166
007 Put the USB flash drive in the storage cabinet. 166
008 Clean the sponge and put it into the basin. 166
009 Throw away the snack packaging, wipe the table. 166
010 Put tableware into the chopstick cage. 166
011 Put fruits into the freshkeeping box. 166
012 Put the mat on the table and then put the cup on the mat. 166
013 Clean the cup with a brush. 166
014 Open the lid and add water to the rice cooker. 166
015 Put the measuring cup and wooden spoon on the tray. 186
016 Wipe the countertop and place the small pot. 166
017 Put the carrot on the board. 166
018 Put the lid on the cup. 166
019 Open the drawer, clip the eraser, and put it in the drawer. 166
020 Open the drawer, clip the stapler, and put it in the drawer. 166
021 Put the ham sausage into the freshkeeping box. 166
022 The spatula goes into the drawer. 166
023 Pick up the object and put it into the sock storage box. 166
024 Pick up objects and put them into the sock storage box. 166
025 One hand picks up the pen and the other hand takes off the cap. 166
026 Pick up the vase and insert flowers. 166
027 Put the items into the small pillow storage box. 166
028 Folding square towels. 166
029 Put baby products into the sterilizer. 166
030 Put the baby shoes into the container. 166
Σ 30 tasks 5,000

📜 Citation

Please consider citing our work if this dataset is useful for your research:

@article{yang2026vista,
  title   = {VISTA: Vision-Grounded and Physics-Validated Adaptation of UMI data for VLA Training}, 
  author  = {Siyuan Yang and Linzheng Guo and Ouyang Lu and Zhaxizhuoma and Daoran Zhang and Xinmiao Wang and Ting Xiao and Fangzheng Yan and Zhijun Chen and Yan Ding and Chao Yu and Chenjia Bai and Xuelong Li},
  journal = {arXiv preprint arXiv:2606.04708},
  year    = {2026},
}